Learning for sensing, localization and connectivity in UAV-aided 6G Networks

The use of terrestrial or flying robots carrying radio equipment is the new promising frontier in our quest towards ever more flexible, adaptable wireless networks. Robot-augmented networks are useful towards extending connectivity, but as it is less known, they also provide unique sensing capabilities. Whether for relaying, mapping or localization purposes, the ultimate selling point of in-network robots lies in their ability to self-navigate in 2D or 3D across meaningful locations so as to improve how data is collected, a concept borrowed from robotics' active learning. This presentation reviews some recent results, illustrated with real-life open-source based experiments, as well as other plentiful scientific and practical challenges for the use of in-network flying robots.